trinity.algorithm.sample_strategy.mix_sample_strategy module

目录

trinity.algorithm.sample_strategy.mix_sample_strategy module#

class trinity.algorithm.sample_strategy.mix_sample_strategy.MixSampleStrategy(buffer_config: BufferConfig, **kwargs)[源代码]#

基类：SampleStrategy

The default sample strategy.

__init__(buffer_config: BufferConfig, **kwargs)[源代码]#

async sample(step: int) → Tuple[List[Experience], Dict, List][源代码]#

Sample data from buffer.

参数:: step (int) -- The step number of current step.
返回:: The sampled List[Experience] data. Dict: Metrics for logging. List: Representative data for logging.
返回类型:: List[Experience]

classmethod default_args() → Dict[源代码]#: Get the default arguments of the sample strategy.

state_dict() → dict[源代码]#: Get the state dict of the sample strategy.

load_state_dict(state_dict: dict) → None[源代码]#: Load the state dict of the sample strategy.